ABSTRACTA near infrared reflectance (NIR) spectrophotometric procedure was evaluated for potential use in assaying nutrient contents of vegetables. Dried ground samples of cauliflower, broccoli, Brussels sprouts, spinach, asparagus, peas, green beans, and squash were analyzed chemically for crude protein, crude fat, ash, and neutral detergent fiber (NDF) and then scanned to develop NIR prediction equations. For the 75 calibration foods samples, the respective correlation coefficients between actual and NIR predicted values of these food components were 0.97, 0.77, 0.91, 0.81. Predictions for 26 unknown samples, which included cabbage, carrots, chard, corn, onions and rhubarb, were based on NIR equations developed from the first 75 samples. The resultant correlations were 0.95, 0.49, 0.85, and 0.78, respectively, for crude protein, crude fat, ash and NDF. NIR prediction detected incorrect chemical values of crude protein in several samples.
展开▼
